Online Influence Maximization under Independent Cascade Model with Semi-Bandit Feedback

We study a stochastic online problem of learning to influence in a social network with semi-bandit feedback, individual observations of how influenced users influence others. Our problem combines challenges of partial monitoring, because the learning agent only observes the influenced portion of the network, and combinatorial bandits, because the cardinality of the feasible set is exponential in the maximum number of influencers. We propose a computationally efficient UCBlike algorithm for solving our problem, IMLinUCB, and analyze it on forests. Our regret bounds are polynomial in all quantities of interest; reflect the structure of the network; and do not depend on inherently large quantities, such as the reciprocal of the minimum probability of being influenced and the cardinality of the action set. To the best of our knowledge, these are the first such results. IMLinUCB permits linear generalization and therefore is suitable for large-scale problems. We evaluate IMLinUCB on several synthetic problems and observe that the regret of IMLinUCB scales as suggested by our upper bounds. A special form of our problem can be viewed as a linear bandit and we match the regret bounds of LinUCB in this case.